Outcome of Very Long-Term Treatment with Antithyroid Drugs in Graves' Hyperthyroidism Associated with Graves' Orbitopathy

Introduction

T here is much debate on the most appropriate way to treat Graves' hyperthyroidism (GH) in the presence of moderate to severe Graves' orbitopathy (GO) (1). Each of the three available treatment modalities has its own advantages and disadvantages in this setting. Treatment with antithyroid drugs appears rather neutral with respect to the course of the eye changes (2), but its disadvantage is the risk on recurrent hyperthyroidism and an associated flare-up of the ophthalmopathy. ¹³¹I therapy is associated with a small but definite risk of worsening of GO; the risk is greater when the eye disease is still active (2 –4). The risk of developing or worsening of eye changes after ¹³¹I therapy in Graves' hyperthyroid patients with no or mild GO can be greatly diminished by a course of prednisone (e.g., 0.25 mg/kg per day for 12 weeks, or even lower doses for a shorter period of time) (3,5). There is, however, no good data on the effectiveness of these rather low doses of prednisone in preventing worsening of eye changes when ¹³¹I therapy is given in patients with moderate to severe and active GO. Much higher doses are likely required, and one could opt for ¹³¹I therapy immediately followed by weekly intravenous methylprednisolone pulses for 3 months. There remains concern about the sharp rise of serum TSH binding inhibitory immunoglobulin (TBII) after ¹³¹I therapy, which is long lasting because low TBII levels are reached only after 5 years (in contrast to treatment with antithyroid drugs or thyroidectomy, where TBII levels start to decline right after initiating therapy and reach low levels in 1.5–2 years) (6). Thyroidectomy was not superior to antithyroid drugs for the eye changes in a case–control study (7). Total thyroidectomy had no advantage over subtotal thyroidectomy for the course of GO in a randomized study (8). Recently, total thyroid ablation (total thyroidectomy followed by radioiodine ablation) was slightly more effective in ameliorating eye changes than total thyroidectomy alone (9), but the degree of improvement was small and lost during follow-up (10). In that study, all patients were also treated with intravenous methylprednisolone pulses. In our opinion, the results of total ablation do not outweigh the disadvantages of such a drastic approach (low but unavoidable risk of surgical complications, and necessity of lifelong treatment with thyroxine).

In our institution, we have long since adopted a peculiar strategy to treat GH in the presence of moderate to severe GO. We prefer a rather conservative approach to the thyroid (thus refraining from radioactive iodine and surgery) and give priority to definitive treatment of the ophthalmopathy in view of the great impact of GO on the quality of life (11). Consequently, we treat hyperthyroidism with antithyroid drugs (30 mg methimazole daily; in case of side effects or pregnancy, it is replaced by propylthiouracil) and choose the block-and-replace (B-R) regimen, which, in our experience, enables us to maintain stable euthyroidism over a long period of time without too many control visits. This policy suits us very well because we discontinue antithyroid drugs only when all medical and/or surgical treatments for GO have been concluded (except prescription of lubricants) and the eye disease does not require any further therapy. This usually takes 2 years or more.

In the present article, we describe the outcome of this strategy. We assessed the recurrence rate of GH after withdrawal of antithyroid drugs, the treatment of recurrences, and whether this affected the course of the ophthalmopathy.

Patients and Methods

In a retrospective follow-up study, we examined consecutive patients referred to the Orbital Center of the Academic Medical Center between 1995 and 2005 because of GO. Inclusion criteria were GO and GH treated according to the B-R regimen for at least 2 years until all medical and/or surgical treatments for GO had been concluded (except prescription of lubricants) and the eye disease did not require any further therapy. Exclusion criteria were treatment with antithyroid drugs according to the titration method, or treatment with ¹³¹I or thyroidectomy before the end of treatment for GO.

Patient files were scrutinized, and questionnaires were sent to subjects who apparently fulfilled the inclusion criteria but did not have an exclusion criterion. The questionnaire contained specific questions on the duration of the B-R regimen and the date when it was discontinued, if they had experienced recurrent GH and how it was treated, and current thyroid medication. Another set of questions related to received medical and surgical treatment for GO, eye changes at the time of discontinuation of antithyroid drugs, and worsening of the ophthalmopathy thereafter until the end of the follow-up; specific attention was called for any temporal relationship of worsening of eye changes and the occurrence of recurrent hyperthyroidism and its treatment. We identified patients with worsening of GO by questions about each specific eye complaint at different time points. Specifically, patients were asked about photophobia, tears, pain behind the eyes, grittiness, red eyes, redness of eyelids, swelling eyelids, exophthalmus, diplopia, disturbed color vision, and/or decreased visual acuity at the time of stopping the B-R regimen and filling in the questionnaire. Also, we asked if they suffered from eye complaints again or if existing complaints had worsened after stopping the B-R regimen, both before and after the treatment of the recurrence of the GH. If they answered yes on any of these latter questions, we considered this as worsening of GO. If the ophthalmopathy had worsened after discontinuation of antithyroid drugs, and also in every case of recurrent hyperthyroidism, we asked the patient's consent to contact their treating physician in order to get independent confirmation of the answers in the questionnaires. The study protocol was approved by the Institutional Review Board of the Academic Medical Center. Written informed consent was returned together with the questionnaire by all subjects finally participating in the study. In case of no response, patients were called asking them again to participate in the study.

Results

The questionnaire was sent to 255 patients selected from the files of the Orbital Center and returned by 114. We had to exclude further 41 patients (10 because of insufficient data, 3 because no informed consent was given, and 28 because the B-R regimen had been changed to another treatment modality by their endocrinologist—most patients were indeed referred back to local specialists for further supervision of antithyroid treatment if thyroid function was stable and thyroid function required only further rehabilitative surgery), leaving 73 patients as the study population (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Number of questionnaires and study population.

The characteristics of the study population, divided into those with and without recurrent GH, at the time of discontinuation of the B-R regimen are given in Table 1. No differences were found between both groups in any of the listed items. The preponderance of female gender, the average age of 49 years when antithyroid drugs were instituted (53 years minus 41 months duration of the B-R regimen), and the 33% prevalence of smoking were all very similar to the characteristics of a representative Dutch population of patients with GH (12). The duration of the B-R regimen was very long, with a median of 3.5 years and a range of 2–11 years. When antithyroid drugs were discontinued, all patients were biochemically euthyroid; TBII was not routinely measured at that time. Almost half of the patients had received steroids and/or orbital irradiation for GO when antithyroid drugs were withdrawn, and rehabilitative orbital surgery also had been done in approximately half of our patients.

Duration of follow-up after stopping the B-R regimen was 4.75 years (range: 1–14 years). The recurrence rate of GH was 37%. Recurrent hyperthyroidism was treated with ¹³¹I in 70% and with antithyroid drugs in 30% (Table 2). Duration of follow-up was not different between the patients with recurrent GH and those remaining in remission. Only two patients indicated in the questionnaire that the ophthalmopathy had worsened after stopping antithyroid drugs. We carefully examined the charts of these two patients and these patients appeared not to have a relapse of GO. Any remaining eye changes did not deteriorate, neither at the time of recurrent hyperthyroidism nor after treatment with ¹³¹I.

Discussion

The first finding of our study is the rate of 37% of recurrent GH, observed after a median 3.5-year course of antithyroid drugs. This recurrence rate might be lower than the usual 40%–50% observed in patients with GH treated with antithyroid drugs (13). One might expect our longer follow-up would result in a higher recurrence rate, but that was not observed. So could the relatively low recurrence rate in the present study be explained from the longer duration of treatment with antithyroid drugs? The Cochrane analysis, however, did not find any relation between the duration of treatment with antithyroid drugs and the recurrence rate (13). But that analysis included studies in which duration of antithyroid drug treatment varies between 6 and 24 months. It is only recently that some studies have been published reporting outcome of treatment longer than 2 years. In general, these uncontrolled observational studies indicate that the longer the duration of treatment with antithyroid drugs, the lower will be the recurrence rate. Retrospective studies showed that long-term treatment with low doses of methimazole is safe (14) and seems to prevent relapse in Graves' disease in patients above 35 years, compared with usual treatment schedule with drug discontinuation (relapse rates of 25% vs. 43%) (15). In our experience, treatment according to the B-R regimen for more than 2 years is not associated with a higher risk of side effects. Moreover, adjustment of the dose in the B-R regimen is seldom necessary and control visits every 6 months are suffice.

The second finding of our study is that—at the time of discontinuation of antithyroid drugs—the characteristics of patients who developed recurrent hyperthyroidism were similar to those who remained euthyroid. The limitation here is the lack of TBII measurements at this point of time, but otherwise thyroid function tests and received treatment for GO were not different between the patients with recurrent GH and the ones without. We also did not measure thyroid volume at this point of time.

The third main finding of our study is that we did not encounter a single case in which the ophthalmopathy worsened after discontinuation of antithyroid drugs. This was also true for the patients in whom recurrent hyperthyroidism was treated with radioactive iodine. According to our policy, antithyroid drugs were discontinued only when all treatment for GO had been completed; patients were thus in the inactive stage of the eye disease, although another limitation of the present study is that the clinical activity score was not always recorded at this time. Our results are in agreement with a previous study reporting no adverse effects of ¹³¹I therapy on ophthalmopathy when delivered in the inactive stage of GO (16). A more recent study, however, observed that GO may occur after radioactive iodine also in the absence of active GO. But they included patients who never had GO before, patients who had inactive GO, and also patients with active GO. Further, it is unclear how their GH was treated and how long antithyroid drugs were given before their GH was treated with radioactive iodine (17).

The low participation rate of our study can be explained by the retrospective nature of our study: patients in whom the eye disease had started 4–14 years earlier were asked to participate. We conclude that our strategy of continuing B-R regimen until GO requires no further treatment is a feasible option. Most encouraging is that late recurrences of hyperthyroidism are not associated with a flare-up of ophthalmopathy. It is very unlikely that, in the absence of a formal eye examination by an ophthalmologist, we missed finding a worsening of eye changes. When asked specifically for each class of eye changes, as was done in this study, patients will very well remember a deterioration as it meant a considerable drawback for them. Also, in favor of our long-term B-R regimen is the suggestion of a lower recurrence rate with longer duration of antithyroid drug treatment. This study raises interesting perspectives on the long-term use of antithyroid drugs in patients with GO, but prospective studies are needed.